Suicide is the third leading cause of death for American adolescents. Nearly one million adolescents a year attempt suicide and about 500,000 adolescents a year are admitted to psychiatric hospitals for suicide attempts or serious suicidal ideation. This leads to high emotional costs for families and financial cost for the health system. Yet, no medication, and less than 10 psychotherapy studies have focused on suicidal youth and findings are mixed. Brent (2006) and Jensen's (2006) call for new and innovative approaches for depression treatment highlights the need for alternative interventions for suicidal youth as well. Attachment-Based Family Therapy (ABFT) offers a promising alternative to prior treatments. It is a manualized family therapy targeting processes associated with suicide and depression. ABFT seeks to improve the adolescent-caregiver relationship by increasing the family's capacity for discussing and negotiating affectively charged issues in the relationship. Improvements in the attachment relationship provide adolescents with improved capacity for affect regulation and the ability to use the caregiver as a source of protection and support. These strengths buffer adolescents against suicide and other risk behaviors. Four studies have demonstrated that ABFT can reduce suicidal ideation and depressive symptoms with an average effect size of .97. Unfortunately, interpretation of these studies is compromised by lack of a controlled comparison treatment. This study aims to test the efficacy of ABFT using a comparison group that controls for treatment dose, duration, therapist expertise, ecological factors, and family involvement. The study includes one year follow-up data, assessment staff blind to treatment condition and tests of the purported active ingredients of ABFT. Putative change processes will be tested including: a) adolescents' expectancies for parent availability, b) emotion regulation during parent-adolescent conflict discussions, and c) resolution of loss and abuse. To test this, Dr. Kobak, a leading adolescent attachment researcher, will use the Adult Attachment Interview and observational coding of the family interaction task to test these treatment mechanisms. If successful, the findings will provide evidence for both the efficacy and specificity of a family based treatment mechanism. We will recruit and randomize 130 adolescents to 16 weeks of ABFT or Family Enhanced-Non-directive Supportive Therapy (FE-NST). Assessments will be conducted at baseline, 8, 16, 32 and 52 weeks. The primary and secondary aims assess whether ABFT reduces suicidal ideation, depression, family conflict, and future suicide attempts more effectively than control. Exploratory aims test a) whether ABFT can improve parent- adolescent attachment, b) if attachment mediates outcome, and if a history of trauma, parental depression or family conflict moderate outcome. The study targets adolescents with severe and persistent suicidal ideation selected from inner city, minority youth.